Ant colony prediction by using sectorized diurnal mobility model for handover management in PCS networks

Recently, mobile phones are extremely used in lifestyle. Historical records of mobile users (MUs) play an important role in predicting future movements of new visitors of the underlying registration area. Handover (handoff) is one of important quality of service (QoS) parameter that affects the continuity of the call when MUs move from a cell to its neighbors in the same registration area (RA). In this paper, a novel ant based Algorithm, has been introduced, which is called Ant Prediction Algorithm (APA). The main target of APA is to reduce handover impact on the performance of personal communication service (PCS) networks. To accomplish such aim, APA tries to minimize the number of dropped calls by predicting the long-term movement of MUs based on the Sectored Diurnal Mobility Model (SDMM). APA consists of two Parts, namely; (i) the Ant Prediction Engine (APE), which relies on the movement history of the other MUs to predict the future movement of the considered MU, and (ii) the SDMM design, which predicts the exact future sector and cell of the considered MU. Simulations have been presented to validate the proposed scheme in terms of prediction accuracy and handoff blocking probability.